Wildly off calculations - what am I getting wrong

[jknorcal]

TL;DR
We are targeting using 18.800KwH/year for a Norcal home, south facing, assuming 5.5 hrs/day sunlight, "extremely complex roof".
My calculations show that I need a 13Kw system with a 125% offset. Tesla's calculations show I need a 18.35Kw system with 103% offset. Who is right?

Detail:
I'm seriously looking into the V3 Tesla roof and have iterated a few times with Tesla on the system size. Their estimate ranged widely from ~10Kw to ~18Kw causing me to questioning the credibility of their numbers.

We are moving into a new home in Northern California. We have 18 months of prior owner's usage. In addition we have 2 electric cars where they didn't have any.

We estimate needing 18,800Kwh/year (1150 Kw/mo x 12 mo + 2 electric cars (at 5000Kwh/year)). We will also get 1 Powerwall.

The home is a two story, south facing, largely away from tall trees. I've been using 5.5 hrs/day sunlight in the calculation. Tesla deem my roof extremely complex.

My estimated system size is 13,000Kw and 1 powerwall. My math looks like this:
(13Kw x 5.5 x 365) x .75 = 19,573Kwh/year (Note: the 25% reduction is something someone suggested - I don't have a clear reason other than a built in offset of 25%)

Tesla is suggesting a 18.35Kw + 1 Powerwall system. In their estimation, this will yield 19,150 Kwh/year with 102% offset. I don't understand how they get this large of a number.

What is the right calculation method?

 

[jjrandorin]

Have you tried putting your roof (and all calculations) into pvwatts website to see if you can get an approximation of what it might generate?

 

[wjgjr]

As @jjrandorin suggested, PVWatts is a great resource to get an estimate of your production. It is worth noting that Tesla tends in many cases to be a bit conservative (perhaps 5-10%, though it is not consistent) with its estimates compared to PVWatts.

The one thing that is missing here is what the actual layout will be. You indicate the roof is south-facing, but, especially with a complex roof, you might not actually be able to fit 13 kW of solar roof on the south-facing side. It is possible that the 18.35 kW number Tesla is quoting is accurate and based on needing to place a good deal of generation on the north-facing side. In terms of kW per area, the solar roof is roughly 2/3 as dense as solar panels (though the actual ratio would depend on the specific roof and local regulations) so it is more common to need to place solar on less ideal portions of the roof in order to reach a desired annual output. For our roof, we ended up maxing it out, with half on the north side. The north-facing side for us generates a bit over half what the south-facing shingles produce.

 

[Vines]

Another factor is that the solar roof tiles once installed simply aren't as efficient as retrofit panels. It depends on the climate, but 10-30% less kWh per kW is what I have heard, though I don't have great data.

Warmer climates where the tile cannot shed the heat as easily lose more, and this makes sense with the lack of airflow under a solar roof compared with retrofit panels.

 

[wjgjr]

Another factor is that the solar roof tiles once installed simply aren't as efficient as retrofit panels. It depends on the climate, but 10-30% less kWh per kW is what I have heard, though I don't have great data.

Warmer climates where the tile cannot shed the heat as easily lose more, and this makes sense with the lack of airflow under a solar roof compared with retrofit panels.

It will be interesting to see actual data on this at some point, including how effective the raised design is at shedding the heat in different climates (and I could imagine things like wind could be factors.)

Nominally, it seems like the solar shingles would only be about 5% less than the panels Tesla is currently installing (including based on data from PVWatts.) As you note, heat may be an additional factor, though, at least in our case, our system performed better during the summer (relative to estimates) than winter. But, as just one data point, it is obviously not enough to draw any conclusion, particularly given the variability of weather out here.

 

[holeydonut]

Another factor is that the solar roof tiles once installed simply aren't as efficient as retrofit panels. It depends on the climate, but 10-30% less kWh per kW is what I have heard, though I don't have great data.

Warmer climates where the tile cannot shed the heat as easily lose more, and this makes sense with the lack of airflow under a solar roof compared with retrofit panels.

I dunno if you saw this post, but the CSI-EPBB calculator only attributes about a 4% production loss for flush/zero-height solar tiles compared to the same system with 6" standoffs:

Standoffs racking can add over 4% vs flush-mount racking

Thanks to @Redhill_qik for posting the link about the CSI-EBPP estimator: http://www.csi-epbb.com/ I guess this tool is a similar thing to PVWatts, and is used by California utilities to do some weird maths around their solar/incentive programs. You can punch in your array spects and it'll...

teslamotorsclub.com

The CSI-EPBB is provided by AESC and the calculator is used by PG&E to determine a reasonable baseline of solar production for a permitted system production. It's basically PVWatts, but with a few more independent variables. PG&E relies on this calculator to determine the monthly amount of solar generation a PTO'd system is capable. Presumably so they can send a QEW to your house and throw a brick through your window if you add more panels after the fact without telling them.

So Tesla's roof may be 30% less efficient than a comparable "normal" system, but the efficiency losses are probably more of an issue of the panels and not their lack of height away from the roof deck.

 

[holeydonut]

Detail:
I'm seriously looking into the V3 Tesla roof and have iterated a few times with Tesla on the system size. Their estimate ranged widely from ~10Kw to ~18Kw causing me to questioning the credibility of their numbers.

IMO, based on the stories I see on TMC and my own experience being a stubborn person trying to "will" my system over the finish line... if you're having issues in the design/spec phase... just cut ties now before you dig yourself into a cess-pit.

The solar corporates are not the type of firms that deal with complexity well. Their processes are all catered to cookie cutter installs with a rigid framework they can push customers through and get a result at the end. Corporates aren't equipped with the customizers and problem solvers to make your special situation work end-to-end.

I highly recommend you find a good, local shop that is equipped with the know-how and flexibility to look at your situation and make things work for you. Yes, the Tesla solar roof V3 looks very nice, but you could be paying for that with sweat/tears and it may not be worth it if your house is difficult. And worse, if you're a discerning customer who values things like offsetting your annual usage and understanding what is happening during the installation... don't expect Tesla Energy to have useful communication at any time during your install.

If your priority is to have a roof that makes your neighbors jealous and to brag about having the best tech on your roof... get the Tesla Solar Roof V3. Throw your giant bag of money at them and get out of Tesla's way.

 

[wjgjr]

IMO, based on the stories I see on TMC and my own experience being a stubborn person trying to "will" my system over the finish line... if you're having issues in the design/spec phase... just cut ties now before you dig yourself into a cess-pit.

The solar corporates are not the type of firms that deal with complexity well. Their processes are all catered to cookie cutter installs with a rigid framework they can push customers through and get a result at the end. Corporates aren't equipped with the customizers and problem solvers to make your special situation work end-to-end.

I highly recommend you find a good, local shop that is equipped with the know-how and flexibility to look at your situation and make things work for you. Yes, the Tesla solar roof V3 looks very nice, but you could be paying for that with sweat/tears and it may not be worth it if your house is difficult. And worse, if you're a discerning customer who values things like offsetting your annual usage and understanding what is happening during the installation... don't expect Tesla Energy to have useful communication at any time during your install.

If your priority is to have a roof that makes your neighbors jealous and to brag about having the best tech on your roof... get the Tesla Solar Roof V3. Throw your giant bag of money at them and get out of Tesla's way.

In fairness, the stories on TMC present a biased picture given that people tend to post when they have issues. And, of course, your issues illustrate how other installers and utilities can also be the issue.

The Tesla solar roof people seem to at least be marginally better than their average, though there is absolutely a major issue with communication. But it should hopefully not take too much work to at least get to a final design and have all the numbers in front of you. With those, you can decide whether to "throw your giant bag of money at them" - or, alternatively, look at how the costs stack up to a re-roofing (if needed) plus panels (noting that non-Tesla panels will likely cost more if the goal is to avoid Tesla,) and what benefits each option brings.

During install, communication will likely be better with a solar roof project as they tend to have larger teams on site for longer, and I would expect that at least the lead will be able to answer questions and address any concerns - certainly that was my experience (though, again, only one experience.)

 

[BGbreeder]

@jknorcal to answer your question, it depends on what your goals are. If it were me, I would go with larger, the Tesla estimate, or more. Your math, as far as I can follow it, seems to assume 100% production from your panels, but you would need to adjust for the fact that your roof isn't at the optimal angle for most or all of the time. I wouldn't put much weight at all in someone else's usage. With two cars, you will be able to utilize a lot of your production, so you may want to go even larger to be able to charge off of solar in the winter time.

As the time based incentives seem to be changing, you may want to consider getting enough Powerwall storage to power your house 4pm to midnight.

If you have the time, model it in PV watts, but "complex" roofs aren't easy.

I guess finally I would say that solar roofs are "new technology", and that if you go that route over "traditional" panels, I would say that you should expect issues (leaks, water missing gutters, wiring failures and inverter issues, and a lack of timely response from Tesla to any issues) that you probably are much less likely to experience with "traditional" panels. Personally, I went with "traditional" panels and microinverters to minimize the potential impact of any failures.

FWIW:
kW = kilowatt, a measure of instantaneous energy (output, total system size)
kWh = kilowatt-hour a measure of energy over time (production/consumption)
MWh= Megawatt hour, a thousand kilowatt hours

All the best,

BG

 

[holeydonut]

In fairness, the stories on TMC present a biased picture given that people tend to post when they have issues. And, of course, your issues illustrate how other installers and utilities can also be the issue.

The Tesla solar roof people seem to at least be marginally better than their average, though there is absolutely a major issue with communication. But it should hopefully not take too much work to at least get to a final design and have all the numbers in front of you. With those, you can decide whether to "throw your giant bag of money at them" - or, alternatively, look at how the costs stack up to a re-roofing (if needed) plus panels (noting that non-Tesla panels will likely cost more if the goal is to avoid Tesla,) and what benefits each option brings.

During install, communication will likely be better with a solar roof project as they tend to have larger teams on site for longer, and I would expect that at least the lead will be able to answer questions and address any concerns - certainly that was my experience (though, again, only one experience.)

I think you're missing the point I made. The Tesla option could be the right option for the right homeowner. My point is the homeowner should avoid getting mired in the details, and needs to just butt out of it so Tesla handle it. But if someone like jknorcal is over here logging into TMC to make an account; going to PVWatts to look at spreadsheets; and otherwise having a strong opinion on what he wants... then I don't think Tesla Energy is right for him.

If some expert at Tesla thinks the "right" system for jknorcal is 18.35 kWp DC + 1x Powerwall... then jknorcal shouldn't start bargaining with Tesla about the details to significantly alter the design. Instead, just sign the contract and look forward to a baller roof.

The more the homeowner wants to get involved to nit pick on designs, usage offsets, DC:AC ratios, shading, weep holes etc... the less likely the resulting experience is going to be as clean/good as yours.

 

[wjgjr]

I think you're missing the point I made. The Tesla option could be the right option for the right homeowner. My point is the homeowner should avoid getting mired in the details, and needs to just butt out of it so Tesla handle it. But if someone like jknorcal is over here logging into TMC to make an account; going to PVWatts to look at spreadsheets; and otherwise having a strong opinion on what he wants... then I don't think Tesla Energy is right for him.

If some expert at Tesla thinks the "right" system for jknorcal is 18.35 kWp DC + 1x Powerwall... then jknorcal shouldn't start bargaining with Tesla about the details to significantly alter the design. Instead, just sign the contract and look forward to a baller roof.

The more the homeowner wants to get involved to nit pick on designs, usage offsets, DC:AC ratios, shading, weep holes etc... the less likely the resulting experience is going to be as clean/good as yours.

I don't want to get into guessing what any individual wants as OP can certainly make up their own mind, but at this point I don't see anybody getting mired into details, just asking some reasonable questions to understand how to judge what they want and would need. My point is that I think it is way to early to be suggesting bailing on Tesla (roof or panels) based on the information we have, and I also disagree with some of the assessments of the rigidity of working with Tesla, particularly on the roof product. For example, if OP says they want a 15 kW roof instead of 18 kW, it is likely Tesla will do that (assuming it does not fall below their minimum threshold for % solar coverage.) But Tesla will tell them that according to their calculations this will only produce x kWh annually, where x is going to be less than the number OP wanted and Tesla based the design on.

(Though I do agree with BGbreeder that larger is likely better, so it is probably worth going with the largest size, as long as cost does not become an issue.)

 

[jknorcal]

@jjrandorin @wjgjr @Vines @holeydonut @BGbreeder You guys are great. Thank you for the input and commentary.
Several suggested scoping it out on pv watts - frankly I'm too dense on this subject and fried after a full day on the computer. I didn't do it.

However I think some of the inputs on this board give me the high order factors to help me get behind Tesla's numbers, namely (in order of significance):

• The surface area of the ideal placement of the EV tiles may be limited to the point of the designer looking to non-ideal areas of the roof to get to the target KwH. And to do that we need more EV tiles for the non-ideal areas than would have needed for the ideal areas for the same Kwh output. (Presumably pv watts would enable me to quantify this) - I'm attaching the tesla roof schematic to illustrate the complexity factor and needing a lot more surface area to get there.
• The efficiency of the solar tiles is lower than solar panels. The sales gal at tesla did say this earlier, I should have asked her to quantify. Numbers vary but certainly a factor
• Trust Tesla - don't nickel and dime it. Ultimately, it just came down to credibility of their numbers not so much me trying to be smarter. Given the above 2 points, clearly I wasn't factoring a couple of items well enough (nor did I use pv watts).

As far as economics, 2 things: 1) the neighborhood and home would carry the investment in the solar roof, 2) I was looking to re-roof anyway away from cedar shake to composite shingle. On 2) I figure, I'm dropping the equivalent of a new tile roof + Panels for this tesla roof.

Thank you folks, I appreciate it.

 

[holeydonut]

@jjrandorin @wjgjr @Vines @holeydonut @BGbreeder You guys are great. Thank you for the input and commentary.
Several suggested scoping it out on pv watts - frankly I'm too dense on this subject and fried after a full day on the computer. I didn't do it.

However I think some of the inputs on this board give me the high order factors to help me get behind Tesla's numbers, namely (in order of significance):

• The surface area of the ideal placement of the EV tiles may be limited to the point of the designer looking to non-ideal areas of the roof to get to the target KwH. And to do that we need more EV tiles for the non-ideal areas than would have needed for the ideal areas for the same Kwh output. (Presumably pv watts would enable me to quantify this) - I'm attaching the tesla roof schematic to illustrate the complexity factor and needing a lot more surface area to get there.
• The efficiency of the solar tiles is lower than solar panels. The sales gal at tesla did say this earlier, I should have asked her to quantify. Numbers vary but certainly a factor
• Trust Tesla - don't nickel and dime it. Ultimately, it just came down to credibility of their numbers not so much me trying to be smarter. Given the above 2 points, clearly I wasn't factoring a couple of items well enough (nor did I use pv watts).

As far as economics, 2 things: 1) the neighborhood and home would carry the investment in the solar roof, 2) I was looking to re-roof anyway away from cedar shake to composite shingle. On 2) I figure, I'm dropping the equivalent of a new tile roof + Panels for this tesla roof.

Thank you folks, I appreciate it.

Sounds like you have a path forward! It sounds like you're in a position to take most advantage of the Tesla solar roof since you need a new roof anyway and have the $ lined up.

I think it's ok to trust Tesla; but you also need patience. The recurring theme for Tesla (Energy and Auto) is they run super lean and act understaffed at almost every turn. I think if you accept this possible reality ... then at least you may have some empathy for the folks that may not get to your question in a timely manner. But for the vast majority of people on TMC who went with Tesla Energy, the wait is worth it in the end .

 

[jknorcal]

Sounds like you have a path forward! It sounds like you're in a position to take most advantage of the Tesla solar roof since you need a new roof anyway and have the $ lined up.

I think it's ok to trust Tesla; but you also need patience. The recurring theme for Tesla (Energy and Auto) is they run super lean and act understaffed at almost every turn. I think if you accept this possible reality ... then at least you may have some empathy for the folks that may not get to your question in a timely manner. But for the vast majority of people on TMC who went with Tesla Energy, the wait is worth it in the end .

 

[DrSmile]

I've had solar/powerwalls/2 electric cars for almost three years now. My lessons learned:

1) You can never have too much solar electricity. I would recommend the maximum amount of panels the utility will let you install because
2) Electric cars need way more power than what you'll estimate. I have a 48 amp/11.5KW charging station and with two cars it's not unusual for the cars to use 60+KWh a day combined, especailly in Winter when your solar production is low and your car consumption is high. I think your 5000 KWh a year number is only realistic if all you do is drive around town a bit each day. Based on my tracking each car uses about 5000 KWh a year.

 

[wjgjr]

@jjrandorin @wjgjr @Vines @holeydonut @BGbreeder You guys are great. Thank you for the input and commentary.
Several suggested scoping it out on pv watts - frankly I'm too dense on this subject and fried after a full day on the computer. I didn't do it.

It seems like you have a good handle on what the numbers are, so no need to use PVWatts - but a good resource to know is available if/as you get curious about the numbers.

It looks like a reasonable layout from Tesla - seems like they pretty much maxed out the South, East, and West roof planes to get to the ~100% offset, which makes sense from a value perspective. You could still choose to oversize, but then it would likely include North-facing panels, so you would end up paying more per kWh of production for that.

 

[h2ofun]

I've had solar/powerwalls/2 electric cars for almost three years now. My lessons learned:

1) You can never have too much solar electricity. I would recommend the maximum amount of panels the utility will let you install because
2) Electric cars need way more power than what you'll estimate. I have a 48 amp/11.5KW charging station and with two cars it's not unusual for the cars to use 60+KWh a day combined, especailly in Winter when your solar production is low and your car consumption is high. I think your 5000 KWh a year number is only realistic if all you do is drive around town a bit each day. Based on my tracking each car uses about 5000 KWh a year.

Totally agree on point 1, which is why I now have about 30KW of solar, 88 panels. No idea how PGE approved, but yep, never can have too much solar!!!!!

 

[h2ofun]

It seems like you have a good handle on what the numbers are, so no need to use PVWatts - but a good resource to know is available if/as you get curious about the numbers.

It looks like a reasonable layout from Tesla - seems like they pretty much maxed out the South, East, and West roof planes to get to the ~100% offset, which makes sense from a value perspective. You could still choose to oversize, but then it would likely include North-facing panels, so you would end up paying more per kWh of production for that.

I have a big set of my panels facing north. It was that or nothing. Yep, not as good as other directions, but better than nothing, and still 26% rebate.